Screen of Vegetation Restoration Patterns in Rocky Desertification Grasslands of Southeast Yunnan Based on Rank-sum Ratio Method

Purpose To address the uncertainty in grass species selection and allocation patterns for vegetation restoration in rocky desertification grasslands of southeast Yunnan, this study conducted experiments with different grass species combinations to screen the optimal vegetation restoration patterns and provide scientific basis for rocky desertification control.

Methods A randomized block design was adopted with six vegetation restoration treatments in mild and moderate rocky desertification areas: monoculture of Lolium perenne (H), monoculture of Dactylis glomerata (Y), monoculture of Setaria sphacelata (G), mixed sowing of D. glomerata and Crotalaria mucronata (Y+S), mixed sowing of S. sphacelata and Trifolium repens (G+B), and mixed sowing of S. sphacelata and Medicago sativa (G+M), with natural vegetation as control (CK). Vegetation characteristics (community coverage, biomass), soil physicochemical properties (nutrient content, moisture status) and other indicators were measured, and comprehensive evaluation was conducted using the rank-sum ratio (RSR) method.

Results 1) All vegetation restoration treatments improved soil physicochemical properties, with soil organic matter, total nitrogen, and available phosphorus increasing by 5.84%-32.62% compared to pre-restoration. 2) Vegetation restoration enhanced community characteristic indicators. In mild and moderate rocky desertification areas, the G+M treatment achieved community coverage of 83.6% and 80.4%, aboveground biomass of 482.0 and 449.5 g/m², and root biomass of 3 002.5 and 2 668.9 g/m², respectively, representing the highest values among all treatments. 3) Water use efficiency differed significantly among treatments, with Poaceae monocultures and Poaceae-Fabaceae mixtures (H, G, G+B, G+M) showing superior performance, ranging from 5.2-6.9 g/(m²·mm) in mild rocky desertification areas and 6.1-7.9 g/(m²·mm) in moderate rocky desertification areas. 4) In mild rocky desertification areas, the G+M treatment had an RSR value of 0.433 (grade: medium), higher than other treatments; followed by G+B and H treatments, with RSR values of 0.383 (grade: medium). In moderate rocky desertification areas, the G+M treatment had an RSR value of 0.933 (grade: excellent), much higher than other treatments; the G+B and H were the secondary optimal treatments, with RSR values of 0.733 and 0.683 (grade: excellent), respectively. The comprehensive evaluation demonstrated that the G+M mixed sowing pattern showed the best performance across different levels of rocky desertification, exhibiting stronger adaptability.

Conclusion In mild and moderate rocky desertification grasslands of southeast Yunnan, mixed sowing of G+M is the optimal vegetation restoration model, and followed by mixed sowing of G+B and H models. The G+M mixed sowing pattern significantly improves vegetation coverage, soil quality, and water use efficiency, and it is recommended as the priority allocation scheme for rocky desertification control in this region.